Table 3 Performance of the AI algorithm for each measurement
From: Evaluation of a deep learning software for automated measurements on full-leg standing radiographs
Lengths and angles | Sample size (N) | RMSE [95% CI] | MAE [95% CI] |
|---|---|---|---|
HKA (°) | 331 | 0.37 [0.34, 0.39] | 0.30 [0.28, 0.32] |
Pelvic obliquity (mm) | 150 | 1.13 [0.87, 1.30] | 0.75 [0.64, 0.86] |
Top leg length (mm) | 309 | 1.45 [1.19, 1.61] | 1.03 [0.92,1.13] |
Center leg length (mm) | 331 | 1.89 [1.68, 2.04] | 1.46 [1.34, 1.57] |
Top femoral length (mm) | 312 | 1.29 [1.14, 1.42] | 0.95 [0.86, 1.04] |
Center femoral length (mm) | 334 | 1.57 [1.44, 1.69] | 1.23 [1.13, 1.32] |
Tibial length (mm) | 344 | 2.02 [1.60, 2.29] | 1.38 [1.21, 1.53] |